and the market

s for the small- Return and Risk

ns for the small-
ks.
. n five types of
uion stocks,
ds, and US.
a •
,
ili u.s.
ent Treasury
bills
,
? 0.1124
5 0.1471
0.1054
0.0880 The previous chapter achieved three pur-
0.0985 poses.First,we acquainted you with the history of U.S.capital markets. Second, we
0.0772 presented statistics such as expected return, variance, standard deviation,
0.0616 covariance,and correlation. Third, we presented a simplified model of the discount
rateon a risky project.
However, we pointed out in the previous chapter that the above model was
nx andy are
adhoc in nature. The next three chapters present a carefully reasoned approach to
negatively
calculatingthe discount rate on a risky project. Chapters 9 and 10 examine the risk
and the return of individual securities, when these securities are part of a large
las for the portfolio.White this investigation is a necessary stepping-stone to discounting
show that if y projects,corporate projects are not considered here. Rather, a treatment of the
appropriate discount rate for capital budgeting is reserved for Chapter 11.
The crux of the current chapter can be summarized as follows: An individual
whoholds one security should use expected return as the measure of the security's
return.Standard deviation or variance is the proper measure of the security's risk.
An individual who holds a diversified portfolio cares about the contribution of
eachsecurity to the expected return and the risk of the portfolio. It turns out that a
securíty'sexpected return is the appropriate measure of the security's contribution
tothe expected return on the portfolio. Thus, our discussion on expected return in
the previous chapter need not be altered, even though portfolios were not ex-
plicitlyconsidered at that time. Conversely, neither the security's variance nor the
security'sstandard deviation is an appropriate measure of a security's contribution
to the risk of a portfolio. Hence, our previous discussions on risk must be greatly
altered to measure a security's contribution to the risk of the entire portfolio of
assets.

Portfolios versus Individual Securities 9.1

Theprevious chapter examined characteristics of individual securities. In particu-
lar,we discussed: 255
256 Part 3 Risk

1. Expected return. This is the return that an individual expects a stock to

earn over the next period. Of course, because this is only an expectation, the actual
return may be either higher or lower. An individual's expectation may simply be Relev
the average return per period a security has earned in the past. Alternatively, it may
be based on a detailed analysis of a firrn's prospects, on some computer-based
model, or on special (or inside) information. Expectations may differ across
individuals, though individuals with similar information are likely to share similar Item
expectations. Expcded return
2. Varianee and standard deviation. There are many ways to assess the Expccted return
volatility of a securíty's return. One of the most common is variance, which is a Varianee of Supe
Varianee of Slow
measure of the squared deviations of a securíty's return from its expected return. Standard deviatic
Standard deviation, which is the square root of the variance, may be thought of asa Standard deviati
standardized version of the variance. Standard deviations and variances for individ- Covariance berw
Correlation berv
ual securities are generally calculated from past data on returns, though more
subjective information may be used as well. A5 we indicated in Figure 8.10,
knowledge of the expected return and the standard deviation can determine the
probability that a securíty's return will fail within a particular range.
3. Covariance and correlation. Returns on individual securities are related
The Expected R
to one another. Covariance is a statistic measuring the interrelationship between
two securities. Alternatively, this relationship can be restated in terms of the The formula for e
correlation between the two securities. The statistical relationship between correla-
The expe
tion and covariance is given in equation (8.2). A5 mentioned earlier, correlation can
oft
be viewed as a standardized version of covariance. Although covariance can take on
any value, positive or negative, correlation can never be greater than + l or less • EXAMPLE
than -l. Consider the two
Suppose that an investor has estimates of the expected returns and standard From the above l
deviations on individual securities and the correlations between securities. How 17.5 percent and
then does the investor choose the best combination or portfolio of securities to The expecn
hold? Obviously, the investor would like a portfolio with a high expected return written as
and a low standard deviation of return. It is therefore worthwhile to consider:
Expecte
1. The relationship between the expected return on individual securities
and the expected return on a portfolio made up of these securities. where Xsuper is
percentage of thi
2. The relationship between the standard deviations of individual se-
Supertech and
curities, the correlations between these securities, and the standard
written as
deviation of a portfolio made up of these securities.
Expected r(
The Example of Supertech and Slowpoke Algebraically, w!
In order to analyze the above two relationships, we will use the same example of
Supertech and Slowpoke that was presented in the previous chapter. The relevant
where XA and X
data from the previous chapter are as follows: 1
respectively. (BJ
equal l or 100 ~
Now cons
expected return
I see Tables 8.3 and 8.4 for actual calculations. regardless of th.
 Return and Risk Chapter 9 257

xpects a stock to
tation,the actual
n may simply be Relevant Data from Example of Supertech
ernatively,it may and Slowpoke
computer-based
ay differ across •
:y to share similar
Itern Symbol Value
Expected return on Supertech !iSuper 0.175 = 17.5%
ays to assess the Expected return on SIowpoke RSlow 0055 = 5.5%
iance, which is a Variance of Supenech
?
(J-Super 0.066875
Variance of Slowpoke 0.013225
,, expected return. (J2Slow
Standard deviation of Supertech (J'Super 0.2586 = 25.86%
be thought of as a Standard devíation of Slowpoke (JSlow 0.1150 = 11.50%
I
;ances for individ- Covariance between Supertech and Slowpoke CTSuper .Slow -0.004875
Correlation between Supertech and Slowpoke -0.1639
ns, though more PSuper .Slow

ti in Figure 8.10,
n determine the
nge.
uritiesare related
tionship between The Expected Return on a Portfolia
in terms of the Theformula for expected return on a portfolia is very simple:
between correla-
ier,correlation can The expected return on a portfolio is simply a weighted average
riancecan take on of the expected returns on the individual securities.
er than + 1 or less • EXAMPLE
Considerthe two securities from the previous chapter, Supertech and Slowpoke.
urns and standard Fromthe above box, we find that the expected returns on the two securities are
n securities. How 17.5percent and 5.5 percent, respectively.
lio of securities to The expected return on a portfolio of these two securities alone can be
hexpected return writtenas
ile to consider:
Expected return on portfolio = Xsupe/17.5%) + Xs'ow(5.5%),
ividualsecurities
I ••
lese secunnes. where Xsuper is the percentage of the portfolio in Supertech and XS10w is the
individualse- percentage of the portfolia in Slowpoke. If the investor with $100 invests $60 in
d the standard Supertech and $40 in Slowpoke, the expected return on the portfolia can be
writtenas
Expected return on portfolio 0.6 x 17.5% + 0.4 x 5.5% 12.7%
Algebraically,we can write
Besame example of Expected return on portfolio = XARA + XBRB (9.1)
apter. The relevant
whereXA and XB are the proportions of the total portfolia in the assets A and B,
respectively.(Because our investor can only invest in two securities, XA + XB must
equal1 or 100 percent.) RA and RB are the expected returns on the two securities.
Now consider two stocks, each with an expected return of 10 percent. The
expectedreturn on a portfolio composed of these two stocks must be 10 percent,
regardlessof the proportions of the two stocks held. This result may seem obvious
258 Part 3 Risk

at this point, but it will become important later. The result implies that you do not
reduce or dissipate your expected return by investing in a number of securities.
Rather, the expected return on your portfolio is simply a weighted average of the
Supertech
expected returns on the individual assets in the portfolio.

Varianee and Standard Deviation of a Portfolio Slowpoke

The Varianee
The formula for the variance of a portfolio composed of two securities, A and B, is

The variance of the portfolio: There a

Var(portfolio) = X~(1~ + 2XAXB(1A/3 + X~(1Í1 obtain equati
The terrn in
Nore that there are three terms on the right-hand side of the equation. The first term in the I
terrn involves the variance of A ((1D, the second term involves the covariance two boxes co
between the two securities ((1AB)' and the third term involves the variance of B cal, indicatin
((1ÍJ). (As we mentioned in Chapter 8, (1A,B = (1B,A' That is, the ordering of the At this
variables is not relevant when expressing the covariance between two securities.) equation (9.2
The formula indicates an important point. The variance of a portfolio de- securities, a tJ
pends on both the variances of the individual securities and the covariance be-
tween the two securities. The variance of a security measures the variability of an Standard D
individual securíty's return. Covariance measures the relationship between the two
Given (9.2'),
securities. For given variances of the individual securities, a positive relationship or return. This i;
covariance between the two securities increases the variance of the entire portfolia.
A negative relationship or covariance between the two securities decreases the
variance of the entíre portfolio. This important result seems to square with corn-
mon sense. If one of your securities tends to go up when the other goes down, or
vice versa, your two securities are offsetting each other. You are achieving what we
!he ínterpre'
cali a hedge in finance, and the risk of your entire portfolio will be low. However, if
ínterpretauor
both your securities rise and fail together, you are not hedging at all. Hence, the
return on ol
risk of your entire portfolio will be higher.
The variance formula for our two securities, Super and Slow, is
15.44%) is 0/
(12.7% + 15
portfolia is ~
Var(portfolio) = XLper(1~uper + 2Xsupe,xSlow(1Super,Slow + X§low(1~low (9.2)
percent occu
Given our earlier assumption that an individual with $100 invests $60 in
Supertech and $40 in Slowpoke, XSuper = 0.6 and XSlow = 0.4. Using this assurnp- The Dioersfic
tion and the relevant data from the box above, the variance of the portfolio is It is ínstructiv
deviation of t
0.023851 = 0.36 x 0.066875 + 2 x (0.6 x 0.4 x (- 0.004875)) + tions of the i
0.16 x 0.013225 (9.2')

The Matrix Approach !There areonlyf,

Alternatively, equation (9.2) can be expressed in the following matrix format: maldistribution
1
 Return and Risk Chapter 9 259

liesthatyou do not Supertech Slowpoke

ber of securities.
ted average of the Supertech
XŠuper<TŠuper X supe~ Slow<TSuper,Slow
0.024075 = 0.36 X 0.066875 - 0.00117 = 0.6 X 0.4 X (- 0.004875)

XSupe~Slow<T Super.Slow XŠlow<TŠlow

Slowpoke
- 0.00117 = 0.6 X 0.4 X (- 0.004875) 0.002116 = 0.16 X 0.013225

.urities,A and B, is

There are four boxes in the matrix. We can add the terms in the boxes to
obtain equation (9.2), the variance of a portfolio composed of the two securities.
B
The term in the upper left-hand corner involves the variance of Supertech. The
terrn in the lower right-hand corner involves the variance of Slowpoke. The other
quation.The first
two boxes contain the term involving the covariance. These two boxes are identi-
'es the covariance
caI, indicating why the covariance term is multiplied by 2 in equation (9.2).
the variance of B
Atthis point, students often find the box approach to be more confusing than
eI ordering of the
equation (9.2). However, the box approach is easily generalized to more than two
tn two securities.)
of a portfolio de- securities, a task we perform later in this chapter.
he covariance be-
e variabilityof an Standard Deviation of a Portfolio
~betweenthe two Given (9.2'), we can now determine the standard deviation of the portfolío's
~erelationship or return. This is
~ entíre portfolia.
"esdecreases the ITp SD (portfolio) VVar(portfolio) = \10.023851
'squarewith corn- 0.1544 = 15.44% (9.3)
I
er goes down, or
~hievingwhat we The interpretation of the standard deviation of the portfolio is the same as the
I
~low.However, if interpretation of the standard deviation of an individual security. The expected
at all.Hence, the return on our portfolio is 12.7 percent. A return of - 2.74 percent (12.7% -
15.44%) is one standard deviation below the mean and a return of 28.14 percent
'w, is (12.7% + 15.44%) is one standard deviation above the mean. If the return on the
portfolio is normally distributed, a return between - 2.74 percent and + 28.14
Ilow(T~low (9.2) percent occurs about 68 percent of the tírne.?
tK) invests $60 in
iSingthis assurnp- The Diversification Effect
e portfolio is It is instructive to compare the standard deviation of the portfolio with the standard
deviation of the individual securities. The weighted average of the standard devia-
)+ tions of the individual securities is
~
D.013225 (9.2')

z There are only four equally probabie returns for Supenech and Slowpoke, so neither security possessesa nor-
mal distribution. TIlUs, probabilities would be slightly different in our example.
aatrixformat:
260 Part 3 Risk

Weighted average of
Suppose
standard deviations = Xsuf>cr uSuf>er + XSlowUSlow I
all the other P,
0.2012 = 0.6 x 0.2586 + 0.4 x 0.115 (9.4) is
One of the most important results in this chapter relates to the difference Variance of d
between (9.3) and (9.4). In our example, the standard deviation of the portfolio is portfolio's ret
less than a weíghted average of the standard deviations of the individual securities.
We pointed out earlier that the expected return on the portfolio is a weighted
average of the expected returns on the individual securities. Thus, we get a The star
different type of result for the standard deviation of a portfolio than we do for the SI
expected return on a portfolio.
It is generally argued that our result for the standard deviation of a portfolio is
due to diversification. For example, Supertech and Slowpoke are slightly negatively Note th
correlated (p = - 0.1639). Supertech's return is likely to be a little below average if portfolio's rei
Slowpoke's return is above average. Similarly, Supertech's return is likely to be a individual rei
little above average if Slowpoke's return is below average. Thus, the standard hence the st
deviation of a portfolio composed of the two securities is less than a weighted below 1. Thi
average of the standard deviations of the two securities.
As long as p
The above example has negatíve correlation. Clearly, there will be less the weigh
benefit from diversification if the two securities exhibited positive correlation. How
high must the positive correlation be before all diversification benefits vanish? In othé
To answer this question, let us rewrite (9.2) in terms of correlation rather perfect corre
than covariance. We mentioned in Chapter 8 that the covariance can be rewritten case of overi
as3 tion. We cou
tion-as lor
USuper,Slow = PSuf>er,Slowu Superu Slow (9.5)
The formula states that the covariance between any two securities is simply the
correlation between the two securities multiplied by the standard deviations of Concept Ql
each. In other words, covariance incorporates both (1) the correlation between the • What are I
two assets and (2) the variability of each of the two securities as measured by deviation
standard deviation.
• What is th
From the previous chapter we know that the correlation between the two
securities is - 0.1639. Given the variances used in equation (9.2'), the standard
deviations are 0.2586 and 0.115 for Supertech and Slowpoke, respectively. Thus,
the variance of a portfolio can be expressed as
The Effi
Our results
Variance of the portfolío's return
In the figur
= '(~uperu~uper + 2XSlIPC~SlowPSuper,slowUsuperUSlow + X;lowU;low dot represe
0.023851 = 0.36 x 0.066875 + 2 x 0.6 x 0.4 x (-0.1639) x security. As
0.2586 x 0.115 + 0.16 x 0.013225 (9.6) standard d
Thet
The middle term on the right-hand side is now written in terms of correlation, p, in Superte.
not covariance.
previously
portfolio.
Thec
3 As with covariance, the ordering of the two securities is not rclevant when expressing the correlation between of an ínfir
the two securities. That is, P~upcr-.Slow = PSIIIW,SUpI ..•
ľ'
sketched l
 Return and Risk Chapter 9 261

Suppose PSuper,Slow = 1, the híghest possible value for correlation. Assume

all the other parameters in the example are the same. The variance of the portfolio
:15 (9.4) is
'the difference Variance of the
~e portfolio is
0.040466 = 0.36 x 0.066875 + 2 x
portfolío's return
ual securities. (0.6 x 0.4 x 1 x 0.2586 x 0.115) + 0.16 x 0.013225
is a weíghted
us, we get a The standard deviation is
we do for the Standard deviation of
VO.040466 = 0.2012 = 20.12% (9.7)
portfolio's return
fa portfolio is
'htly negatively ote that (9.7) and (9.4) are equal. That is, the standard deviation of a
Iowaverage if portfolío's return is equal to the weighted average of the standard deviations of the
likely to be a individual returns when P = 1. lnspection of (9.6) indicates that the variance and
í, the standard
hence the standard deviation of the portfolio must drop as the correlation drops
an a weighted below 1. This leads to:
As long as P < 1, the standard deviation of a portfolio of two securities is less than
e will be less the weighted average of the standard deviations of the individual securities.
rrelation. How
iefits vanish? In other words, the diversification effect applies as long as there is less than
Irelation rather perfect correlation (as long as P < 1). Thus, our Supertech-Slowpoke example is a
n be rewritten case of overkill. We illustrated diversification by an example with negative correla-
tion. We could have íllusrrated diversification by an example with positive correla-
tion-as long as it was not perfect positive correlation.
(9.S)
is simply the
deviations of Concept Questions
on between the • What are the formu las for the expected return, variance, and standard
measured by deviation of a portfolio of two assets?
• What is the diversification effect?
etween the two
), the standard
ectívely, Thus,
The Effident Set for Two Assets '.2
Our results on expected returns and standard deviations are graphed in Figure 9.1.
In the figure, there is a dot labelled Slowpoke and a dot labelled Supertech. Each
+ X~low(J~low dot represents both the expected return and standard deviation for an individual
security. As can be seen, Supertech has both a higher expected return and a higher
.013225 (9.6) standard deviation .
The box or "O" in the graph represents a portfolio with 60 percent invested
f correlation, p, in Supertech and 40 percent invested in Slowpoke. You will recall that we have
previously calculated both the expected return and the standard deviation for this
portfolio.
The choice of 60 percent in Supertech and 40 percent in Slowpoke is just one
correlationbetween of an infinite number of portfolios that can be created. The set of portfolios is
sketched by the curved line in Figure 9.2.
 262 Pan 3 Risk

Figure 9.1 Expected return (%)

straight line
Expected return and coefficient b
standard deviation
trated in the
for (l) Supertech,
(2) Slowpoke, and Considerpo
Supertech
17.5 and 10 pero
(3) a ponfolio com-
posed of 60 percent • argue that ti
in Supertech and 40
percent in effect applie
Slowpoke. 12.7 o point l' but
the clutter (
Thouj
9.2, they de
5.5 •
Slowpoke the curve e
investor ca:
can not cho
Standard 2. T
11.50 15.44 25.86 deviation (%)
portfolio VI
have the lc
Figure 9.2 folio is sta
Expected return
Set of ponfolios of portfolio (%) standard d
composed of hold-
an ce, is rm
ings in Supertech
and Slowpoke (cor- 3. j
relation between the Supertech and Super
two securities is 17.5 line in Fíg
-0.16).
approprial
Portfólio J is com-
the curve
posed of 90 percent
Slowpoke and 10 decrease
percent Supertech between t
(p = -0.16). MV
he canno
Portfólio 2 is com-
posed of 50 percent deviation~
5.5
Slowpoke and 50 Slowpoke want to a
percent Supertech Wei
(p = -0.16)
Portfolío 3 is com- could eVI
I I I Standard
posed of 10 percent 11.50 25.86 deviation
investor'
Slowpoke and 90 little risk
of portfolio's
percent Supertech
(p = - 0.16). return (%) minimun
Ponfolio ľ is com- 4.
posed of 90 percent and MV.
Slowpoke and 10 Consider portfolio 1. This is a portfolio composed of 90 percent Slowpoke
percent Supertech actually (
and 10 percent Supertech. Because it is weighted so heavily toward Slowpoke, it
(p = l). can an
Point MV denotes
appears close to the Slowpoke point on the graph. Portfolio 2 is higher on the
curve because it is composed of 50 percent Slowpoke and 50 percent Supertech. reductio
the minimum vari-
ance portfolio. This Th
Portfolio 3 is dose to the Supertech point on the graph because it is composed of
is the portfolio with
90 percent Supertech and 10 percent Slowpoke. twO seci
the lowest possible up wher
variance. By defini- There are a few important points concerning this graph.
tion, the same of Supei
1. We argued that the diversification effect occurs whenever the correlation
portfolío must also the port
have the lowest pos-
between the two securities is below 1. The correlation between Supertech and
always (
sible standard Slowpoke is - 0.1639. This diversification effect can be illustrated by comparison
deviation. with the straíght line between the Supertech point and the Slowpoke point. The bends t
 Return and Risk Chapter 9 263

straightline represents points that would have been generated had the correlation
coefficient between the two securities been 1. The diversification effect is illus-
trated in the figure since the curved line is always to the left of the straight line.
Considerpoint 1'. This represents a portfolio composed of 90 percent in Slowpoke
and10 percent in Supertech if the correlation between the two were exactly 1. We
argue that there is no diversification effect if p = 1. However, the diversification
effectapplies to the curved line, because point 1 has the same expected return as
point l' but has a lower standard deviation. (Points 2' and 3' are omitted to reduce
the clutter of Figure 9.2.)
Though the straíght line and the curved line are both represented in Figure
9.2,they do not simultaneously exist in the same world. Either p = -0.1639 and
the curve exists or p = 1 and the straight line exists. In other words, though an
investor can choose between different points on the curve if p = - 0.1639, she
cannot choose between points on the curve and points on the straight line.
ird
lon(%) 2. The point MV represents the minimum variance portfolio. This is the
portfolio with the lowest possible variance. By definition, this portfolio must also
havethe lowest possible standard deviation. (The term minimum varianee port-
folia is standard in the literature, and we will use that termo Perhaps minimum
standard deviation would actually be better, because standard deviation, not vari-
ance,is measured on the horizontal axis of Figure 9.2.)
3. An individual contemplating an investment in a portfolio of Slowpoke
and Supertech faces an opportunity set or feasible set represented by the curved
line in Figure 9.2. That is, he can achieve any point on the curve by selectíng the
appropriate mix between the two securities. He cannot achieve any points above
the curve because he cannot increase the return on the individual securities,
decrease the standard deviations of the securities, or decrease the correlation
betweenthe two securities. Neither can he achieve points below the curve because
he cannot lower the returns on the individual securities, increase the standard
deviationsof the securities, or increase the correlation. (Of course, he would not
wantto achieve points below the curve, even if he were able to do so.)
Were he relatively tolerant of risk, he might choose portfolio 3. (In fact, he
could even choose the end point by investing all his money in Supertech.) An
rd
In investorwith less tolerance for risk might choose point 2. An investor wanting as
ollo's littlerisk as possible would choose MV,the portfolio with minimum variance or
%)
minimumstandard deviation.
4. Note that the curve is backward bending between the Slowpoke point
ent Slowpoke and MV.This indicates that, for a portion of the feasible set, standard deviation
tl Slowpoke, it actuallydecreases as one increases expected return. Students frequently ask, "How
higher on the can an increase in the proportion of the risky security, Supertech, lead to a
nt Supertech. reduction in the risk of the portfolio?"
; composed of This surprising finding is due to the diversification effect. The returns on the
twosecurities are negatively correlated with each other. One security tends to go
up when the other goes down and vice versa. Thus, an addition of a small amount

~~~~:~~~~~~
ly companson
ofSupertech acts as a hedge to a portfolio composed only of Slowpoke. The risk of
the portfolio is reduced, implying backward bending. Actually, backward bending
alwaysoccurs if p :s; O. It mayor may not occur when p > O. Of course, the curve
'(kepoint. The bends backward only for a portion of its length. As one continues to increase the
 264
Pan 3 Risk

Figure 9.3
Expected return
Opponunity sets packages
ofportfolio
composed of hold-
ings in Supenech preferre
p =- I
and Slowpoke. decision
P = - 0.5
allocate
Each curve repre- p=O
sents a different
portfolio
correlation. The An
lower the correla- folios th
tion, the more bend
in the curve.
portfólio
returns,
the peri
portfolio
portfolid
small p
reduces
L Standard words, t
deviation than olli
of portfolio's
return

percentage of Supenech in the portfolio, the high standard deviati on of this

security eventually causes the standard deviation of the entire portfolio to rise.
5. No investor would want to hold a portfolio with an expected return
below that of the minimum variance portfolio. For example, no investor would
choose portfolio 1. This portfolio has less expected return but more standard
deviation than the minimum variance portfolio has. We say that portfolios such as
portfolio 1 are dominated by the minimum variance portfolio. Though the entire
curve from Slowpoke to Supenech is called the feasible set, investors only con sider
the curve from MV to Supenech. Hence, the curve from MV to Supenech is calied
the efficíent set.

Figure 9.2 represents the opponunity set where p = -0.1639. It is worth-

Total
while to examine Figure 9.3, which shows differem curves for different correla- of porU
tions. As can be seen, the lower the correlation, the more bend there is in the 0.2
curve. This indicates that the diversification effect rises as p declines. The greatest 0.19
bend occurs in the limiting case where p = - l. This is perfect negatíve correla- 0.18
tion. While this extreme case where p = - l seems to fascinate students, it has little 0.17
practicai imponance. Most pairs of securities exhibit positive correlation. Strong 0.16
negative correlation, let alone perfect negative correlation, are unlikely OCCur- 0.15
rences tndeed.s
0.14
The graphs we examined are not mere intellectual curiosities. Rather, effi- 0.13
cíent sets can easily be calculated in the real world. As mentioned earlier, data on 0.12
rerurns, standard deviations, and correlations are generally taken from past data, 0.11
though subjective notions can be used to calculate the values of these statistics as O.I
well. Once the statistics have been determined, any one of a whole host of software 0.09
0.08
0.07
0.06
4 A major exception occurs with derivative securities. For example, the correlation between a stock and a put on O
the stock is generally strongly negative. Puts will be treated later in the text.
 Return and Risk Chapter 9 265

Standard
deviati on
of portfolio's
return
lard devíatíon of this
e portfolia to rise.
an expected return
e, no investor would
but more standard
at portfolias such as
lo. Though the entire
I •
\vestorsonly consíder
10 Supertech is called
packages can be purchased to generate an efficient set. However, the choice of the
preferred portfolia within the effícíent set is up to you. As with other important
decisions like what job to choose, what house or car to buy, and how much time to
allocate to this course, there is no computer program to choose the preferred
portfolia.
An efficient set can be generated where the two individual assets are port-
folios themselves. For example, the two assets in Figure 9.4 are a diversified
portfolia of American stocks and a diversified portfolio of foreign stocks. Expected
returns, standard deviations, and the correlation coefficient were calculated over
the period from 1973 to 1988. No subjectivity entered the analysis. The U.S. stock
portfolia with a standard deviation of about 0.173 is less risky than the foreign stock
portfolia, which has a standard deviation of about 0.222. However, combining a
small percentage of the foreign stock portfolio with the ns. portfolio actually
reduces risk, as can be seen by the backward-bending nature of the curve. In other
words, the diversification benefits from combining two different portfolios more
than offset the introduction of a riskier set of stocks into one's holdings. The
minimum variance portfolio occurs with about 80 percent of one's funds in
American stocks and about 20 percent in foreign stocks. Addition of foreign
securities beyond this point increases the risk of one's entire portfolio.
The backward-bending curve in Figure 9.4 is important information that has
not bypassed American money managers. In recent years, pensíon-fund and mu-
tual-fund managers in the United States have sought out investment opportunities
overseas. Another point worth pondering concerns the potential pitfalls of using
only past data to estimate future returns. The stock markets of many foreign
countries such as Japan have had phenomenal growth in past years. Thus, a graph
like Figure 9.4 makes a large investment in these foreign markets seem attractive.
However, because abnormally high returns can not be sustaíned forever, some
subjectivity rnust be used when forecasting future expected returns.

0.1639. It is worth- Tolal return Figure 9.4

of portfolio (%) Retuml risk tradeoff
for different correla-
0.2 for world stocks.
end there is in the portfolia of u.s. and
0.19 .
M tntmum 0%U5 . .. 100 % F oretgn
clines. The greatest . foreign srocks,
ect negative correla-
0.18 vananee ------a
0.17 ~..........-- 10%
students, it has little portfolio ~---
0.16
correlation. Strong .......--0 40 Ol
30
•
cl. 20 % U 5
ib ... 80 % Forei
0.15
are unlikely occur- 0.14 ,cl / 50% ' ',"

0.13 60%
osities. Rather, effí- 0.12 70%
ed earlier, data on 0.11
en from past data, O.I
f these statistics as 0.09
ble host of software 0.08

0.07 [ Risk (standard

0.06 ' , , , , , , 'deviation of
0.15 0.17 0.19 0.21 0.23 portfolio's
n a stock and a PU! on return) (%)
266 Pan 3 Risk

Concept Question
• What is the relationship between the shape of the efficient set for two assets
and the correlation between the two assets?

'.l The Efficient Set for Many Securities

The previous discussion concerned two securities. We found that a simple curve
sketched out all the possible portfolios. Because investors generally hold more
than two securities, we should look at the same curve when more than two
securities are held. The shaded area in Figure 9.5 represents the opportunity set or
feasible set when many securities are considered. The shaded area represents all
the possible combinations of expected return and standard deviation for a port-
folio. For example, in a universe of 100 securities, point 1 might represent a
portfolio of, say, 40 securities. Point 2 might represent a portfolio of 80 securities.
Point 3 might represent a dífferent set of 80 securities, or the same 80 securities
held in different proportíons, or something else. Obviously, the combinations are
virtually endless. However, note that all possible combinations fit into a confined
region. No security or combínatíon of securities can fail outside of the shaded
region. That is, no one can choose a portfolio with an expected return above that subje
given by the shaded region because the expected returns on individual securities 10001
cannot be altered. Furthermore, no one can choose a portfolio with a standard corre
deviation below that given in the shady area. Perhaps more surprisingly, no one can of see
choose an expected return below that given in the curve. In other words, the
capital markets actually prevent a self-destructíve person from taking on a guaran- derivt
teed loss.> prino
So far, Figure 9.5 is different from the earlier graphs. When only two se- softw
curities are involved, all the combinations lie on a single curve. Conversely, the size.
combinations cover an entire area with many securities. However, notice that an woul
individual will want to be somewhere on the upper edge between MV and X. The
upper edge, which we indicate in Figure 9.5 by a thick line, is called the efficient
set. Any point below the efficient set would receive less expected return and the Varil
same standard deviation as a point on the efficient set. For example, consider Ron Weť.
the efficient set and W directly below it. lf W contains the risk you desire, you asset
should choose R instead in order to receive a higher expected return. wort
In the final analysis, Figure 9.5 is quite similar to Figure 9.2. The efficient set man
in Figure 9.2 runs from MV to Supertech. It contains various combinations of the view
securities Supertech and Slowpoke. The efficient set in Figure 9.5 runs from MV to
X. It contains various combinations of many securities. The fact that a whole shaded
the 1
area appears in Figure 9.5 but not in Figure 9.2 is just not an important difference; asse
no investor would choose any point in Figure 9.5 below the efficient set anyway. on t
We mentioned before that an efficient set for two securities can be traced out
easily in the real world. The task becomes more difficult when additional securities vert
are included because the number of observations grows. For example, using peľ<

s Of course, someone dead set on paning with his money can do so. For examplc, he can trade frcqucntly without
purpose, so that commissions more than offset the positive expected returns on the portfólio. 6111e
 Return and Risk Chapter 9 267

Expected return Figure 9.5

of portfolio The feasible set of
for two assets portfolios con-
x structed from many
securilies.

t a simple curve
erally hold more
more than two MV
pportunity set or
tea represents all
iation for a port-
"ght represent a
of 80 securities.
I Standard
e 80 securities deviation
ombinations are of portfolio's
into a confined return

~t~:nt~~osv~a:~
subjectiveanalysis to estimate expected returns and standard deviations for, say,
vidual securities
100 or 500 securities may very well become overwhelming, and the difficulties with
twitha standard
correlationsmay be greater still. There are almost 5,000 correlations between pairs
ngly,no one can
of securities from a universe of 100 securities.
ther words, the
Though much of the mathematics of efficient-set computation had been
ng on a guaran-
derivedin the 1950s,6 the high cost of computer time restricted application of the
principles. In recent years, the cost has been drastically reduced. A number of
in only two se-
softwarepackages allow the calculation of an efficient set for portfolios of moderate
Conversely, the
size.Byall accounts these packages sell quite briskly, so that our above discussion
, notice that an
wouldappear to be important in practice.
MV and X. The
ed the effícíent
return and the Varianee and Standard Deviation in a Portfolio of Many Assets
consider R on
Weearlier calculated the formulas for varianee and standard deviation in the two-
bu desire, you
assetcase. Because we considered a portfolio of rnany assets in Figure 9.5, it is
um.
worthwhile to calculate the formulas for variance and standard deviation in the
he efficient set

,~:~~n~~~
whole shaded
many-assetcase. The formula for the varianee of a portfolio of many assets can be
viewedas an extension of the formula for the variance of two assets.
To develop the formula, we employ the same type of matrix that we used in
the two-asset case. This matrix is displayed in Table 9.1. Assuming that there are N
ant difference;
assets,we write the numbers 1 through N on the horizontal axis and 1 through N
t set anyway.
on the vertical axis. This creates a matrix of N x N = NZ boxes.
be traced out
Consider, for example, the box with a horizontal dimension of 2 and a
nalsecurities
verticaldimension of 3. The term in the box is X3 x, Cov(R3' Rz). X3 and x, are the
cample, using
percentages of the entire portfolio that are invested in the third asset and the

frequently without
6 The classic is I larry Markowitz, Portfolio Selec/ion (New York: John Wiley & Sons, 1959).
268 Pan 3 Risk

stocks.
Table 9.1 Matrix used to calculate the varianee of a portfolio
bers of
Stock 2 3 N diagon
stocks i
)(far x,XlCov(RI> Rl) x,X3Cov(R" R3) X,XrvCov(R, , RN)
terms)
2 XzX1Cov(Rz, R,) Xžaž XzX3Cov(Rl, R3) XzXrvCov(Rl, RN) portfoli
folío's I
3 X~,Cov(R3, R,) X~lCov(R3, Rl) X3a5 X~rvCov(R3, RN)

N XNX1Cov(RN> R,) XNXlCov(RM Rl) X~3Cov(RN> R3) X2~F,

Conce
cr, is the standard deviation of stock i.
Cov(R" R) is the covariance between stock i and stock J. • What
Terms involving the standard deviation of a single security appear on the diagonal. Terms involving covariance • How
between two securities appear off the díagonal.

second asset, respectively. For example, if an individual with a portfolio of $1,000 Dive
ínvests $100 in the second asset, X2 = 10% ($100/$1,000). Cov(R3, R2) is the
The al
covariance between the returns on the third asset and the returns on the second
Suppc
asset. Next, note the box with a horízontal dimension of 3 and a vertical dimension
of 2. The term in the box is X2 X3 Cov(R2, R3)' Because Cov(R3, R2) = Cov(R2, R3),
both boxes have the same value. The second security and the third security make
up one pair of stocks. In fact, every pair of stocks appears twice in the table, once in
the lower left-hand side and once in the upper right-hand side.
Suppose that the vertical dimension equals the horizontal dimension. For
example, the term in the box is Xiai when both dimensions are one. Here, a~ is
the variance of the return on the first security.
Thus, the diagonal terms in the matrix contain the variances of the different

Table 9.2 Number of varianee and covariance terms as a function simpl

of the number of stocks in the portfolio all of
portf
Number of Number of
varianee covariance diagc
terms terms terffii
umber of TOlaI (number of (number of
stocks in number terms terms expľl
portfólio of terms on diagonal) off diagonal)

1 1 1 O
2 4 2 2
3 9 3 6
10 100 10 90
100 10,000 100 9,900

N Nl N N2 - N

In a large portfolio, the number of terms involving covariance between two securities
is much greater than the number of terms involving variance of a single security.
 Return and Risk Chapter 9 269

stocks.The off-diagonal terms contain the covariances. Table 9.2 relates the nurn-
bersof diagonal and off-diagonal elements to the size of the matrix. The number of
N diagonalterms (number of variance terms) is always the same as the number of
stocksin the portfolio. The number of off-díagonal terms (number of covariance
X,XtvCov(R" RN)
terms) rises much faster than the number of diagonal terms. For example, a
X?XtvCov(R2, RN) portfolio of 100 stocks has 9,900 covariance terms. Since the variance of a port-
folío's returns is the sum of all the boxes, we have:
XyXtvCov(R3, RN)
The variance of the return on a portfolio with many securities
is more dependent on the covariances between the individual
securities than on the variances of the individual securities.
I
r XZ,pF,
[ Concept Questions
• What is the formula for the variance of a portfolio for manyassets?
IS involving covariance • How can the formula be expressed in terms of a box or matrix?

nfolío of $1,000 Diversification:AnExample 9.4

~0v(R3' R2) is the
The above point can be illustrated by altering the matrix in Table 9.1 slightly.
mson the second
Suppose that we make the following three assumptions:
ertícaldimension
lz) = Cov(R2' R3), 1. All securities possess the same variance, which we write as var. In other
lrd security make words, 0'1 = var for every security.
thetable, once in 2. All covariances in Table 9.1 are the same. We represent this uniform
covariance as cov. In other words, Cov(R;, R) = cov for every pair of
il dimension. For securities. It can easily be shown that var> cov.
e one. Here, O''f is 3. All securities are equally weighted in the portfolio. Because there are N
assets, the weight of each asset in the portfolio is 1/N. In other words,
es of the different
X; = 1/N for each security i.

Table 9.3 is the matrix of variances and covariances under these three
~nction simplifyingassumptions. Note that all of the diagonal terms are identicaI. Similarly,
all of the off-diagonal terms are identicaI. As with Table 9.1, the variance of the
Imber of
portfoliois the sum of the terms in the boxes inTable 9.3. We know that there are N
rianee diagonalterms involving variance. Similarly, there are N x (N - l) off-díagonal
errns
mber of
terms involving covariance. Summing across all the boxes in Table 9.3, we can
erms express the variances of the portfolio as
lííagonal)
r----
Yariance
o of N x ~2 )var + N(N - l) x (~2)COV
2
6 portfolio
90 Number of Each Number of Each
~,900 diagonal diagonal off-díagonal off-diagonal
terms term terms term

2-N
(l)'VY var +
(N2 N2- N) cOV
ecuruies
rity, (~ ) var + (1 - ~) cov (9.8)
270 Part 3 Risk

--
Table 9.3 Matrix used to calculate the variance of a
portfolio when (a) all securities possess the
same variance, which we represent as var; (b)
all pairs of securities possess the same
covariance, which we represent as cov, (c) all
securities are held in the same proportion,
which is liN.

Stock 2 3 N

(l/NZ)Var (l/NZ)cov (l/NZ)Cov (l/NZ)Cov

2 (l/NZ)Cov (l/NZ)var (l/NZ)Cov (l/NZ)Cov

3 (l/NZ)Cov (l/NZ)cov (l/NZ)var (l/NZ)Cov

N (l/NZ)Cov (l/NZ)cov (l/NZ)cov (l/NZ)var

Suppc
Equation (9.8) expresses the variance of our special portfolio as a weighted sum of time c
the average security variance and the average covaríance.? The intuition appears
perce
when we increase the number of securities in the portfolio without limit. The
origir
variance of the portfolio becomes
Variance of portfolio (when N ~ oo) = cov (9.9) Fígun
becai
This occurs because (1) the weight on the variance term, liN, goes to O asN goes to The v
infinity, and (2) the weight on the covariance term, 1 -liN, goes to 1 as N goes to
of th
infinity. portf
Formula (9.9) provides an interesting and important result. In our special thec
portfolio, the variances of the individual securities completely vanish as the
number of securities becomes large. However, the covariance terms remain. In
addit
fact, the variance of the portfolio becomes the average covariance, cov. One often síons
l hears that one should diversify. You should not put all your eggs in one basket. The
too l
effect of diversification on the risk of a portfolio can be illustrated in this example. worl
The variances of the individual securities are diversified away, but the covariance
singl
terms can not be díversífiedaway.
buyi
~ The fact that part, but ňOt all, of one's risk can be diversified away should be díve
exp lo red. Consider Mr. Smith, who brings $1,000 to the roulette table at a casino. It
achu
would be very risky if he put all his money on one spin of the wheel. For example,
imagine that he put the full $1,000 on red at the table. If the wheel showed red, he
would get $2,000, but if the wheel showed black, he would lose everything.
8 This

9 The
covar
7 Equation (9.8) is actually a weighted aoerage of the varianee and eovarianee terms because the weights, lIN and 10 Me
1 - lIN, sum tO l.
(Sept.
 271
Return and Risk Chapter 9

Figure 9.6
Relationship be-
Variance
tween the vanance
of
of a portfolio's re-
portfolio's
turn and the
return
number of securities
Diversifiahle risk, in the portfolio-*
unique risk, or
unsystematic risk * This graph
assumes
VAR a. AII securities
have constant
\ variance, var.
b. AII securities
have constant
covariance, rov.
COV
!l'------------------ Portfolio risk,
market risk, or
c. AII securities are
equallly
weighted in
systematic risk portfolio.
Number of The variance of a
securities portfolio drops as
2 3 4 more securities are
added tO the port-
folio. However, it
does not dro~
Suppose,instead, he divided his money over 1,000 different spíns by betting $1 at a zero. Rather, cov
serves as the floor.
[ghted sum of time on red. probability theo ry tells us that he could count on winning about 50
ition appears percent of the time. In other words, he could count on pretty nearly getting all his
:ut limit. The original $1,000 back.8
Now, let's contrast this with our stock-market example, which we íllustrate in
Figure9.6. The variance of the portfolio with only one security is, of course, var
(9.9) becausethe variance of a portfolio with one security is the variance of the security.
pasNgoes to Thevariance of the portfolio drops as more securities are added, which is evidence
~ asN goes to of the diversification effect. However, unlike Mr. Smith's roulette example, the
portfolío'svariance can never drop to zero. Rather it reaches a floor ofčôv, which is

I
n our special the covariance of each pair of securities?
<,
Because the variance of the portfolio asymptotically approaches cov, each
ľanish as the
s remain. In additional security continues to reduce risk. Thus, if the re were neither commis-
lOV. One often
sions nor other transactions costs, it could be argued that one can never achieve
re basket. The too much diversification. However, the re is a cost to diversification in the real
.this example. world. Commissions per dollar invested fail as one makes larger purchases in a
he covariance single stock. Unfortunately, one must buy fewer shares of each security when
buying more and more dífferent securities. Comparing the costs and benefits of
wayshould be diversification,Meir Statman argues that a portfolio of about 30 stocks is needed to
e at a casino. It achieveoptimal diversification.10
..For example,
owed red, he
e everything.
8 This ignores the casíno's cut.
• Though it is harder tO show, this risk reduction effect also applies to the general case where variances and

covariances are not equal.

10 Meir Statman, "How Many Stocks Make a Diversified ponfolio'" jouY1UlI of Fi/Ulncial and Quantitative Analysis
he weights, liN and
(Seplembcr 1987). I
272 Pan 3 Risk

We mentioned ear1ier that var must be greater than cov. Thus, the variance of gambles
a securíty's return can be broken down in the following way: advanta~

Total risk of Unsystematic or

individual security = Portfolio risk + diversifiable risk
(Var) (cov) (var - cov) Concept

Total risk, which is var in our example, is the risk that one bears by holding onto • What a
one security only. Portfolio risk is the risk that one still bears after achieving full • Why d,
diversification, which is cov in our example. Portfolio risk is often called systematic
or market risk as well. Diversifiable, unique, or unsystematic risk is that risk that
can be diversified away in a large portfolio, which must be (var - cov) by RisIde
definition. Figure 9
To an individual who selects a diversified portfolio, the total risk of an an ínves
individual security is not important. When considering adding a security to a or rísk-f
diversified portfolio, the individual cares about that portion of the risk of a security íllustrate
that can not be diversified away. This risk can alternatively be viewed as the
contribution of a security to the risk of an entire portfolio. We will talk later about • EXAM
the case where securities make different contributions to the risk of the entire Ms. Logi
portfolio. addition
cifuV paramet

Risk and the Sensible Investor

Having gone to all this trouble to show that unsystematíc risk disappears in a well-
diversified portfolio, how do we know that investors even want such portfolios?
Suppose they like risk and don 'r want it to disappear?
We must admit that, theoretically at least, this is possible, but we will argue
that it does not describe what we think of as the typical investor. Our typical
investor is risk averse. Rísk-averse behavior can be defined in many ways, but we Suppos
prefer the following example: A fair gamble is one with zero expected return; a ínveste
rísk-averse investor would prefer to avoid fair gambles. pected
Why do investors choose well-diversified portfolios? Our answer is that they returns
are risk averse, and rísk-averse people avoid unnecessary risk, such as the un-
Expe
systematic risk on a stock. If you do not think this is much of an answer to why
COI
investors choose well-díversified portfolios and avoid unsystematic risk, consider
whether you would take on such a risk. For example, suppose you had worked all
summer and had sayed $5,000, which you intended to use for your college Becaus
expenses. Now, suppose someone came up to you and offered to flip a eoin for the return
money: heads, you would double your money, and tails, you would lose it all. calcula
Would you take such a bet? Perhaps you would, but most people would not. equatic
Leaving aside any moral question that might surround gambling and recognízíng l
that some people would take such a bet, it's our view that the average investor writter
would not.
To induce the typical risk-averse investor to take a fair gamble, you must
x1
sweeten the pot. For example, you might need to raise the odds of winning from Howex
50-50 to 70-30 or higher. The rísk-averse investor can be induced to take fair (TMervil
 Return and Risk Chapter 9 273

I the variance of gambles only if they are sweetened so that they become unfair to the investor's
advantage.
c or
lrisk
w) Concept Questions
y holding onto • What are the two components of the total risk of a security?
- achieving full • Why doesn't diversification elirninate all risk?
led systematic
is that risk that
if - cov) by Riskless Borrowing and Leading ,.s
Figure 9.5 assumes that all the securities on the efficient set are risky. Alternatively,
tal risk of an an investor could easily combine a risky investment with an investment in a ~ss
security to a or risk-free security, such as an investment in United States Treasury bills. This is
k of a security illustrated in the following example.
\iewed as the
tik later about • EXAMPLE
of the entire Ms. Logue is considering investing in the common stock of Merville Enterprises. In
addítíon, Ms. Logue will either borrow or lend at the risk-free rate. The relevant
parameters are

Common stock Risk-free

of Merville asset
ars in a well-
th portfolios? Expected Guaranteed
return: return:
Return 14% 10%
ve will argue
Standard deviation: 0.20 O
Our typical
ways, but we Suppose Ms. Logue chooses to invest a total of $1,000, $350 of which is to be
!ed return; a ínvested in Merville Enterprises and $650 placed in the risk-free asset. The ex-
pected return on her total investment is simply a weighted average of the two
r is that they returns:
I as the un-
Expected return on portfolio
rwer to why
I
Sk, consider
composed of one riskless 0.114 = 0.35 x 0.14 + 0.65 x 0.10 (9.10)
and one risky asset
worked aJl
ur college Because the expected return on the portfolio is a weíghted average of the expected
oin for the return on the risky asset (Merville Enterprises) and the risk-free return, the
se it all. calculation is analogous to the way we treated two risky assets. In other words,
would not. equatíon (9.1) applies here.
ecognizing Using equation (9.2), the formula for the variance of the portfolio can be
ge investor written as

you must X~erville(J~erville + 2X MervillcX Risk-free(J Merville,Risk-free + Xfusk-free(Jfusk-free

ning from However, by definition, the risk-free asset has no variability. Thus, both
o take fair and (J~isk-free are equal to zero, reducing the above expression to
ITMerville.Risk-free
274
Pan 3 Risk

Figure 9.7
Expected return
Relationship be-
on portfolio (%)
tween expected
return and risk for a
ponfolio of one
120% in Merville Corp.
risky asset and one
20% in risk-free asset s
riskless asset.
(borrowing at risk-free
rate)

14
J

10=RF Borrowing to invest in

MerviJJe Corp. when the
borrowing rate is greater
35% in MerviJJe Corp.
than the lending rate
65% in risk-free asset

l , Standard
20 deviation
of portfolio's
return (%)

Variance of portfolio composed

of one riskless and one risky asset
X~erviJJe(J~erviJJe = (0.35)2 X (0.20)2 0.0049 (9.11)
The standard deviation of the portfolio is
Standard deviation of portfolio composed
of one riskless and one risky asset =
XMerviJJe(JMerviJJe = 0.35 x 0.20 = 0.07
(9.12)
The relationship between risk and return for one risky and one riskless asset
can be seen in Figure 9.7. Ms.Logue's split of 35-65 percent between the two assets
is represented on a straight line between the risk-free rate and a pure investment in
Merville Corp. Note that, unlike the case of two risky assets, the opponunity set is
straíghr, not curved.
Suppose that, alternatively, Ms. Logue borrows $200 at the risk-free rate.
Combining this with her original sum of $1,000, she invests a total of $1,200 in the
Merville Corp. Her expected return would be
Expected return on portfolio
formed by borrowing = 14.8% = 1.20 x 0.14 + (-0.2) X 0.10
to invest in risky asset

Here, she Invests 120 percent of his original investment of $1,000 by borrowing 20
percent of her original investment. Nore that the return of 14.8 percent is greater
than the 14-percent expected return on Merville Corp. This occurs because she is
borrowing at 10 percent to invest in a security with an expected return greater than
10 percent.
 Return and Risk Chapter 9 275

The standard deviation is

Standard deviation of portfolio

formed by borrowing 0.24 1.20 x 0.2
to invest in risky asset

The standard deviation of 0.24 is greater than 0.20, the standard deviation of the
Merville Corp., because borrowing increases the variability of the investment. This
investment also appears in Figure 9.7.
So far, we have assumed that Ms. Logue is able to borrow at the same rate at
which she can lend.!! Now let us consider the case where the borrowing rate is
above the lending rate. The dotted line in Figure 9.7 illustrates the opportunity set
for borrowing opportunities in this case. The dotted line is below the solid line
because a higher borrowing rate lowers the expected return on the investment.

The Optimal Portfolia

The previous section concerned a portfolio formed between one riskless asset and
one risky asset. In reality, an investor is likely to combine an investment in the
riskless asset with a portfolio of risky assets. This is illustrated in Figure 9.8.
Consider point Q, representing a portfolío of securities. Point Q is in the
interior of the feasible set of risky securities. Let us assume the point represents a
portfolio of 30 percent in AT&T, 45 percent in General Motors (GM) and 25 percent
in IBM. Individuals combining investments in Q with investments in the riskless
(9.11 )
asset would achieve points along the straíght line from RF to Q. We refer to this as
line f. For example, point 1 represents a portfolio of 70 percent in the riskless asset
and 30 percent in stocks represented by Q. An investor with $100 choosing point 1
as his portfolio would put $70 in the risk-free asset and $30 in Q. This can be
restated as $70 in the riskless asset, $9 (0.3 x $30) in AT&T, $13.50 (0.45 x $30) in
(9.12) GM, and $7.50 (0.25 x $30) in IBM. Point 2 also represents a portfolio of the risk-
asset free asset and Q, with more (65%) being invested in Q.
assets Point 3 is obtained by borrowing to invest in Q. For example, an investor with
aent in $100 of his own would borrow $40 from the bank or broker in order to invest $140
set is in Q. This can be stated as borrowing $40 and contributing $100 of one's own
money in order to invest $42 (0.3 x $140) in AT&T, $63 (0.45 x $140) in GM, and
e rate. $35 (0.25 x $140) in IBM.
in the Though any investor can obtain any point on line l, no point on the line is
optima!. To see this, consider line II, a line running from RF through A. Point A
represents a portfolio of risky securities. Line II represents portfolios formed by
combinations of the risk-free asset and the securities in A. Points between RF and A
10 are portfolios in which some money is invested in the risk less asset and the rest is
placed in A. Points past A are achieved by borrowing at the riskless rate to buy more 1.

ing 20 of A than one could with one's original funds alone.

,reater
she is
" Surprisingly, this appears to be a decent approximation because a large number of investors are able to borrow
h than on margin when purchasing stocks. The borrowing rate on the margin is very near the rískless rate of interest,
particularly for large investors. More will he said about this in a larer chapter.
276 Pan 3 Risk

Figure 9.8 Expected return

Relationship be- of portfolio Line (capital-market line)
1/ portf
tween expected
return and standard in th
deviation for an ín- line
vestment in a rate
combination of risky
securities and the She
riskless asset. that i.
inter
Portfólio Q is corn-
posed of
30 percent AT&T
Risk-free
rate ( RF)
14 l
{2
l·
lc:
7
C
+
Line
I

45 percent GM
25 percent IBM 35% in risk-free asset - 40% in risk-free asset
65% in stock s 140% in stocks
70% in risk-free asset represented by Q represented by Q
30% in stocks
represented by Q
Standard
deviation
of portfolio's
return

drawn, line /l is tangent to the efficient set of risky securities. Whatever

As
Deti!
point an individual can obtain on line l, he can obtain a point with the same TheJ
standard deviation and a higher expected return on line II. In fact, because line II is and I

tangent to the efficient set, it provides the investor with the best possible oppor- . curit]
tunities. In other words, line Il, which is frequently called the capital market line, estl~
• I

can be viewed as the efficient set of all assets, both risky and riskless. An investor
with a fair degree of risk aversion might choose a point between RF and A, perhaps
• be~al
price
point 4. An individual with less risk aversion might choose a point closer to A or
even beyond A. For example, point 5 corresponds to an individual borrowing
money to increase his investment in A.
saJ
neve
The graph illustrates an important point. With riskless borrowing and lend- worl
ing, the portfolio of risky assets held by any investor would always be point A. tion
L- r Regardless of the ínvestor's tolerance for risk, he would never choose any other
point on the efficient set of risky assets (represented by curve XAY) nor any point in sam
the interior of the feasible region. Rather, he would combine the securities of A
with the riskless assets if he had high aversion to risk. He would borrow the riskless ~!trJ
I
asset to invest more funds in A had he Iowaversion to risk. wou
This result establishes what financial economists cali the separation principie. assd I
That is, the investor makes two separate decísions:
1. After estimating Ca) the expected return and variances of individual the
securities, and (b) the covariances between pairs of securities, the investor calcu- aver
lates the efficient set of risky assets, represented by curve XAY in Figure 9.8, and 4, fc
determines Point A, the tangency between the risk-free rate and the efficient set of SE
risky assets C curve XAY). Point A represents the portfolio of risky assets that the
investor will hold. This point is deterrnined solely from her estímates of returns,
variances, and covariances. No personal characteristics, such as degree of risk- 12 Th
aversion, are needed in this step. varia!
 Return and Risk Chapter 9 277

line)
2. The investor must now determine how she will combine point A, her
portfolioof risky assets, with the riskless asset. She could invest some of her funds
in the riskless asset and some in portfolio A. She would end up at a point on the
linebetween RF and A in this case. Alternatively, she could borrow at the risk-free
rateand contribute some of her own fund as well, investing the sum in portfolio A.
Shewould end up at a point on line /l beyond A. Her position in the riskless asset,
that is, her choice of where on the line she wants to be, is determined by her
internalcharacteristics, such as her ability to telerate risk.

Concept Questions
• Whatis the formula for the standard deviation of a portfolio composed of one
risklessand one risky asset?
• Howdoes one determine the optimal portfolio among the efficíent set of risky
assets?
's

Market Equilibrium 9.6

. Whatever Definition of the Market-Equilibrium Portfolio
the same Theabove analysis concerns one investor. Her estimates of the expected returns
e line II is and variances for individual securities and the covariances between pairs of se-
ble oppor- . curitiesare hers and hers alone. Other investors would obviously have dífferent
~arket line,
I estimatesof the above variables. However, the estimates might not vary much
\rl investor
A,perhaps
becauseall investors would be forming expectations from the same data on past •
pricemovement and other publicly available information.
ser to A or Financial economists often imagine a world where all investors .B0~sessthe
borrowing same estírnates on expected returns, variances, and covariances. Though this can
neverbe literally true, it can be thoug~ of as a useful simplifying assumption in a
and lend- worldwhere investors have acce~ to similar sources of information. This assump- ~
point A. tion is called homogeneous expectations. J 2
any other If all investors have homogeneous expectations, Figure 9.8 would be the
iy point in samefor all individuals. That is, all investors would sketch out the same efficíenr set
ities of A ofriskyassets because they would be working with the same inputs. This effícíent
tieriskless setof risky assets is represented by the curve XAY Because the same risk-free rate
I <
wouldapply to everyone, all investors would view point A as the portfolio of risky
vrinciple. assetsto be held.
This point A takes on great importance because all investors would purchase
ndividual the riskysecurities that it represents. Those investors with a high degree of risk-
or calcu- aversionmight combine A with an investment in the riskless asset, achieving point
9.8, and 4,forexample. Others with Iowaversion to risk might borrow to achieve, say, point
nt set of 5. Becausethis is a very important conclusion, we restate it:
! that the
f returns,
~ of risk-
" The assumption of homogeneous expectations states that all investors have the same beliefs concerning returns,
variances,and covariances. It does not say that all investors have the same aversion to risk.
278 Part 3 Risk

In a world with homogeneous expectations, all investors would hold

the portfolio of risky assets represented by point A. Table 9.5

If all investors choose the same portfolio of risky assets, it is possible to

determine what that portfolio is. We show below that it is the market portfolio .

• EXAMPLE
Imagine a stock market composed of only four securities: Alpha Electronics, Beta
Food Products, Gamma Clothing, and Delta Home Builders. Financial detail s on the
four securities are provided in Table 9.4.
Further imagine that there are only three investors holding stocks. All of these
Alpha Elect
investors have homogeneous expectations. The investors and their wealth in stocks
are Beta Food

Ownership Gamma Ch
Wealth in of market
Delta Horn
stock market (percent)

Mary Russell $50,000,000 50 Total

Vincent Dinoso 49,980,000 49.98
Walter Peck 20,000 0.02 • 3 million

Total $100,000,000 100 t $15 millio

Under the assumption that all three individuals have homogeneous expecta-
tions, we know from our earlier analysis that all investors must be holding identical
portfolios of risky assets. Suppose that each investor holds shares in a security
based on the percentage of the market he or she owns. That is This
Shares Percentage Number ual is hok
of a security = of the entire x of shares security ir
held by an market that the of security (9.13) ~ investors.
individual individual owns outstanding shows th.
outstandn
For example, Walter Peck owns 0.02% of the market. Thus, he would hold the mark,
600 shares (0.02% x 3 million shares) of Alpha Electronics Alte
SOOshares (0.02% x 2.5 million shares) of Beta Food Products portfolio
investor
míned bt
thus the:
Table 9.4 Financial detaiIs of securities in the market place
of Delta l
(l) (2) (3) (4) (5) price, th
Price Number of Total* Percentage
shares
not deal
er market of
Name R
s are outstanding value market fall until
market t
$15,000,000
Alpha Electronics $ 5 3,000,000 $15,000,000 15%
$100,000,000
portfolic
= $5 x 3,000,000
Beta Food Products $10 2,500,000 $25,000,000 25 Ir
Gamma Clothing $20 2,000,000 $40,000,000 40
Delta Home Builders $40 500,000 $20,000,000 20
$100,000,000 100

• TOtaI marker value = Price per share X Number of shares outstandíng. 13 Perhaps
 Return and Risk Chapter 9 279

ould hold
A Table 9.5 Holdings of each investor in dífferent stocks assuming that each investor takes a position in a
stock proportionate to the number of the stock's outstanding shares
it is possible to
arket portfolio. Mary Russell Vincent Dinoso Walter Peck Total

(1) (2) (3) (4) (5) (6) (7) (8)

' Dollar
Electronics, Beta value of
cialdetaiIs on the Shares held stock held
Dollar Dollar Dollar by all by all
Shares value Shares value Shares value investors investors
locks.All of these
A1pha Electronics 1,500,000 $7.5 1,499,400 $ 7.497 600 $3,000 3 $15
wealth in stocks
mill ion million million* million]
Beta Food Products 1,250,000 $12.5 1,249,500 $12.495 500 $5,000 2.5 $25
million million million million
Gamma Clothing 1,000,000 $20 999,600 $19.992 400 $8,000 2 $40
million million million million
Delta Home Builders 250,000 $10 249,900 $ 9.996 100 $4,000 0.5 $20
mill ion million million mill ion
-- --- --
Total $50 $49.98 $20,000 $100
million million million

• 3 million = 1,500,000 + 1,499,400 + 600

t Sl5 million = $7.5 million + $7.497 million + $3,000

eneous expecta-
lolding ídentícal 400 shares (0.02% X 2 million shares) of Gamma Clothing
les in a security 100 shares (0.02% X 0.5 million shares) of Delta Home Builders

This strategy is often called holding the market portfolia because the individ-
ual is holding a constant percentage of the number of outstanding shares in each
security in the marketplace.P Table 9.5 represents this strategy for each of the three
(9.13) investors. A comparison of column (7) in Table 9.5 with column (3) of Table 9.4
shows that under this strategy investors in aggregate hold exactly the number of
outstanding shares of each security. In other wor-ís, financial economists say that
uld hold
the market clears. >?
Alternatively, suppose that all investors want to hold the same non-market
ucts portfolio. For example, suppose that, because Alpha is selling at a low price, each
investor wants to hold more shares of Alpha Electronics than the number deter-
mined by formula (9.13). Investors want more shares of Alpha than are outstanding,
• ,,;
thus the market does not clear. Similarly if each investor wants to hold fewer shares
ofDelta Home Builders than determined by (9.13) because Delta is selling at a high
(S)
/,
price, they want fewer shares of Delta than are outstanding, and the market does
Percentage
of not clear. In equilibrium, the price of Alpha must rise and the price of Delta must
market fail until investors willingly hold the exact number of outstanding shares. The
~ $15,000,000
market then clears. This would occur only when each investor holds the market
I $100,000,000 portfolio. Thus, we have a very important result:
In a world of homogeneous expectations, the market will only clear.
when each investor holds the market portfolio.
II~

13 Perhaps it WOli Id be more correct to say each investor hold, a portion or percemage of the market portfolío.
280 Part 3 Risk

Definition of Risk When Investors Hold the Market Portfolio securi

in a lar,1
The previous section shows that all individuals rnust hold the market portfólio in a
world with homogeneous expectations. This result allows us to be more precise is equa
about the risk of a security in the context of a diversified portfolio. We represent
We arg'
the variance of a portfolio as the sum of all the boxes in Table 9.6. Table 9.6 is portfolf
identical to Table 9.1 except for shadings to emphasize security 2. the rnai
Consider security 2 in the matrix. What is the contribution of this security to inside
the risk of the entire portfolio? We can answer this question by examining the security
shaded row in Table 9.6. As indicated by the shaded area, security 2 appears in each the mal
box in the row. The row contains the covariances between security 2 and all of the It
securities in the portfolío, including the covariance of security 2 with itself. (The
covariance of security 2 with itself, Cov(R2, R2), is Var(R2) == (J~ by definition.) That Cov(R21
is, the row can be written as
where
= X2XICov(R2, Rl) + X~COV(R2' R2) + X2X3Cov(R2, R3)
return
+ ... + X2XNCov(R2, RN)
the risl
= X2[XICov(R2, Rl) + X2Cov(R2, R2) + X3Cov(R2, R3) return
+ ... + XNCov(R2, RN)] nothini
= X [ContribUtiOn of security 2 to the risk of the entire portfolio, ] (914) market
2 standardized by the percentage of security 2 in the portfolio . sented

One can argue that (9.14) measures the contribution of security 2 to the risk of the
portfolio. F
The second row of (9.14) factors out X2, the percentage of security 2 in the to bel
entire portfolio. Clearly, the contribution of security 2 to the risk of the entire
portfolio is proportional to X2, the percentage of security 2 in the entire portfolio.
Thus, the terms in (9.14) within brackets can be viewed as the contribution of
security 2 to the risk of the entíre portfolio, standardized by the percentage of
14 The e

few cruc
semed a
Table 9.6 Matrix used to calculate the varianee of a portfolia

Stock 2 3 N where o
surns aCI
XT<Tr XIX2Cov(RI, R2) XIX3Cov(RI> R3) XIX~ov(RI> RN)
change
2 X~ICov(R2, Rl) X1<T~ X~3Cov(R2, R3) X~~ov(R2, RN)

3 X~ICov(R3, RI) X~2Cov(R3' R2) X1<T3 X~~ov(R3, RN)

The ten

111e 2 il

. volved .

N X~lCov(RN, Rl) X~2Cov(RM R2) X~3Cov(RN, R3) X'fpF,

IS By ne
This chart is idemical la that in Table 9.1 except that we are speaking now of a specific portfolio---the market
portfolia. X, now stands for the proportion of security i in the market portfolio.
 Return and Risk Chapter 9 281

~lio
~et portfolio in a
pe more precise
10. We represent
19.6. Table 9.6 is
~ portfolio. Under this assumption of homogeneous
~fthis security to
examining the
r appears in each
2 and all of the
ith itself. (The
~
~efinition.) That
security 2 in the portfolio. This is perhaps the best measure of the risk of a security
in a large portfolio.!?
The set of terms in (9.14) within brackets has an interesting interpretation. X;
isequal to the proportion of each individual's portfolio that is invested in security i. ?
We argued that, under homogeneous expectations, all investors hold the market
expectations, X; is the ratio of
the market value of security i to the value of the entire market. Thus, the terms
inside the bracket represent a weighted average of the covariances between
security 2 and each security in the market, where the weights are proportional to
the market value of each security in the market.
It can be shown that
Cov(R2, RM) = XICov(R2, Rl) + X2Cov(R2, R2) + X3Cov(R2, R3)
+ . . . + XN Cov(R2, RN)

where Cov(R2, RM) is the covariance between the return on security 2 and the
return on the market portfolio. Thus, the standardized contribution of security 2 to
!XNCov(R2, RN)
the risk of the market portfolio is proportional to the covariance between the
return on security 2 and the return on the market as a whole. Of course, there is
lNCOV(R 2, RN)] nothing special about security 2. The contribution of any security i to the risk of the
r]
D
(9.14)
market portfolio, standardized by its percentage in the portfolio, can be repre-
sented as

l the risk of the

Researchers have further standardized
Cov(R;, RM)

(9.15) by defining the beta of security i

(9.15)
f ~/
~curity 2 in the to be'>
k of the entire
~tire portfolio. Cov(R;, RM)
13; (9.16)
(I2(RM)
uuon o f
Iontrtibutí
percentage of

" The explanation we have presented is heuristíc in nature. We confess that there has been a sleíght of hand at a
few crucial points. A more rigorous derivation can be stated as follows. The variance of the portfolío can be repre-
sented as
N N
CIJ, = Variance of portfolío = J, j~, X,Xpij <a)
N
where CIij is the covariance of; with} if i '*}. and CIiI is the variance or CIT if i = j. The double surnrnation in (a)
I sums across all boxes in Table 9.6. •
rM=Ov(Rj, RN) The contribution of security i to the risk of a portfólio can be best wrítten as oCI;.loX,. TI,;s measures the
change in the variance of the entire portfólio when the proportíon of security; is increased slightly. For security 2,
~Ov(R2' RN)
cJ(J~ N ..,
oj = 2 L Xp'2 = 2[X,Cov(R" R2) + X2CIi + X3Cov(R3, RJ + ... + XN Cov(RN' R2)1 (b)
2 j= l
'M=OV(R3, RN)
The term within brackets in (b) is Cov(R2, RM)' Hence we can rewrite (b) as

~Z
I
= 2 Cov(R2, RM) (c)

The 2 in (c) occurs because the terms involving security 2 in both the second row and the second column are in-
volved. Though the varianee term, 0'22 = (T~, occurs only once, nore that

)(2tf1Ft --ax;--
OX'CI'
=
2x z(J 22

io-the market " By noung that Cov(R" RM) = P'é,PPM' we can rewrite J3,as

J3, = P'M5!!...
, (J",

I "
282 Part 3 Risk

where (T2(RM) is the variance of the market. Though both Cov(Ri> RM) and l3i can be
used as measures of the contribution of security i to the risk of the market portfolia,
l3i is much more common. One useful property is that the average beta across all
securities, when weighted by the proportion of each security's market value to that
of the market portfolio, is 1. That is,
'\.
N

LXi13i
i= 1
1 (9.17)

Beta as a Measure of Responsiveness

The previous discussion shows that the beta of a security is the standardized
covariance between the return on the security and the return on the market.
Though this explanation is 100% correct, it is not likely to be 100% intuitively
appealing to anyone other than a statístícían. Luckily, there is a more intuitive
explanation for beta. We present this explanation through an example .
• EXAMPLE
Consider the following possible returns on both the stock of jelco, Inc., and on the
market:
Return on Return on
Type of market jelco, Inc.
State economy (percent) (percent)

l Bull 15 25
II Bull 15 15
re
III Bear -5 -5
tv
N Bear -5 -15
l~
Though the return on the market has only two possible outcomes (15% and - 5%), P
the return on jelco has four possible outcomes. It is helpful to consider the
expected return on a security for a given return on the market. Assuming each state cl
is equally likely, we have 11

p
Return on Expected return
Type of market on jelco, Inc. a
economy (percent) (percent) p
c
Bull 15% 20% = 25% x Y2 + 15% X 1/2
Bear -5% -10% = -5% x Y2 + (-15%) X Y2

jelco, Inc., responds to market movements because its expected return is greater in
bullish states than in bearish states. We now caleulate exactly how responsive the
security is to market movements. The markeťs return in a bullish economy is 20
percent (15% - (- 5%» greater than the markeťs return in a bearish economy.
However, the expected return on jelco in a bullish economy is 30 percent (20% -
(-10%» greater than its expected return in a bearish state. Thus.jelco, Inc., has a
responsiveness coefficient of 1.5 (30%/20%).
This relationship appears in Figure 9.9. The returns for both jelco and the
market in each state are plotted as four points. In addition, we plot the expected
 Return and Risk Chapter 9 283

Id 13;can be Figure 9.9

Return on security (%)
Performance of
et portfolio, jelco, Inc., and the
Characteristic line
6 across all
25
market portfolia.
lue to that 20 (20%,15%)*

(9.17) 10

-15 15 25

indardízed
e market. - 10
intuitively
·e intuitive
- 20
Return on
market (%)
nd on the • The two points marked X represent the expected return on jelco for each possible outcome
of the market portfolio. The expected return on jelco is positively related to the return on
the market. Because the slope is 1.5, we say that jelco's beta is 1.5. Beta measures the
responsiveness of the securiry's return LO movement in the market.
• (20%, 15%) refers to the point where the return on the security is 20% and the return on
the market is 15%.

return on the security for each of the two possible returns on the market. These
two points, which we designate by X, are joined by a line caIled the characteristic
line of the security. The slope of the line is 1.5, the number calculated in the
d -5%), previous paragraph. This responsiveness coefficient of 1.5 is the beta of jelco.
isíder the The interpretation of beta from Figure 9.9 is intuitive. The graph teIls us that
ach state the returns on jelco are magnified 1.5 times over th ose of the market. When the
market does well.jelco's stock is expected to do even better. When the market does
poorly.Ielco's stock is expected to do even worse. Now imagine an individual with
a portfolia near that of the market who is considering the addition of jelco to his
portfolia. Because of jelco's magnificationfactor of 1.5, he will view this stock as
contributing much to the risk of the portfolia. We showed earlier that the beta of
lh
the average security in the market is 1. jelco contributes more to the risk of a large,
) x '12
diversified portfolia than does an average security because jelco is more respon-
greater in sive to movements in the market.
Insive the Further insight can be gleaned by examining securities with negative betas.
rmy is 20 One should view these securities as either hedges or insurance policies. The
economy. security is expected to do well when the market does poorly and vice-versa.
t(20% - Because of this, adding a negative beta security to a large, diversified portfolio
~c., has a actuallyreduces the risk of the portfolia. 16

t and the
expected lb untonunarely, empirical evidence shows that Vit11131ly no st()ck~ have negative beta".
284 Part 3 Risk

A Test
,
We have put this question on past corporate finance examínatíons:

1. What sort of investor rationally views the variance (or standard devia-
tion) of an individual security's return as the security's proper measure
of risk?
2. What sort of investor rationally views the beta of a security as the se-
curity's proper measure of risk?
A proper answer might be something like the following:

A rational, risk-averse investor views the variance (or standard deviation) of her
portfolío's return as the proper measure of the risk of her portfolio. lf for some
reason or another the investor can hold only one security, the variance of that se-
curiry's return becomes the variance of the portfolío's return. Hence, the varianee
of the securiry's return is the security's proper measure of risk. the eí
lf an individual holds a díversified portfolío, she still views the variance (or íllustri
standard deviation) of her portfolío's return as the proper measure of the risk of secud
her portfolio. However, she is no longer interested in the variance of each individ-
ual securíty's return. Rather, she is interested in the contribution of an individual
security to the variance of the portfolio.
risk-In
Under the assumption of homogeneous expectations, all individuals hold the return
market portfolio. Thus, we measure risk as the contribution of an individual
security to the variance of the market portfolio. This contribution, when standard- securi
ized properly, is the beta of the security. While very few investors hold the market of the
portfolio exactly, many hold reasonably diversified portfolios. These portfolios are each ~
~
close enough to the market portfolio so that the beta of a security is likely to be a secun
reasonable measure of its risk. for an

cause
expeq
,
Concept Questions
• If all investors have homogeneous expectations, what portfolio of risky assets som~
do they hold? pecte

• What is the formula for beta?

secur
• Why is beta the appropriate measure of risk for a single security in a large secur
portfolio? by bi
secu~
SML.
9.7 Relationship between Risk and Return high<
It is commonplace to argue that the expected return on a security should be
positively related to its risk. That is, individuals will hold a risky security only if its beloi
expected return compensates for its risk. This reasoning holds regardless of the borr
measure of risk. Now consider our world where all individuals (1) have homoge- SML,
neous expectations and (2) all individuals can borrow and lend at the risk-free rate.
All individuals hold the market portfolio of risky securities here. We have shown price
that the beta of a security is the appropriate measure of risk in this context. Hence, adju]
 Return and Risk Chapter 9 285

Expected return Fígure 9.10

ns. on security (%) Relationship be-
Security market line (SML) tween expected
return on an individ-
ndard devia- ual security and beta
\roper measure of the security.
RM
RF is the risk-free
ity as the se- rate.
RM is the expeaed
RF return on the mar-
ket portfolia.

tíon) of her
. lf for Some Beta of
ince of that se- o 0.8 security
e, the variance
the expected return on a security should be positively related to its beta. This is
variance (or
ílíustrated in Figure 9.10. The upward-sloping line in the figure is called the
of the risk of
security-market line (SML).
f each individ-
an individual There are six important points associated with this figu re.
1. A beta of zero. The expected return on a security with a beta of zero is the
risk-freerate, Rp. Because a security with zero beta has no relevant risk, its expected
fuals hold the return should equal the risk-free rate.
n individual 2. A beta of one Equation (9.17) points out that the average beta across all
en standard- securities, when weighted by the proportion of each securíry's market value to that
Id the market of the market portfolio, is 1. Because the market portfolio is formed by weighting
ľonfolios are eachsecurity by its market value, the beta of the market portfolio is 1. Because all
likelyto be a securities with the same beta have the same expected return, the expected return
forany security with a beta of 1 is RM' the expected return on the market portfolio.
3. Linearity. The intuition behind !n upwardly sloping curve is clear. Be-
cause beta is the appropriate ineasure of risk, high-beta securities should have an
expected return above that of low-beta securities. However, Figure 9.10 shows
!sky assets something more than an upwardly sloping curve, the relationship between ex-
pected return and beta corresponds to a straigbt line.
It is easy to show that the line in Figure 9.10 is straight. To see this, consider
a large security S with, say, a beta of 0.8. This security is represented by a point below the
security-market line in the figu re. Any investor could duplicate the beta of security S
by buying a portfolio with 20 percent in the risk-free asset and 80 percent in a
securitywith a beta of 1. However, the homemade portfolio would ítself lie on the
SML. In other words, the portfolio dominates security S because the portfolio has a
higher expected return and the same beta.
hould be ow consider security T with, say, a beta greater than 1. This security is also
only if its belowthe SML in Figure 9.10. Any investor could duplicate the beta of security T by
less of the borrowing to invest in a security with a beta of 1. This portfolio must also lie on the
• homoge- SML, thereby dominating security T
-free rate. Because no one would hold either S or T, their stock prices would drop. This
e shown priceadjustment would raise the expected returns on the two securities. The price
t. Hence, adjustment would continue until the two securities lay on the security-market line.
286 Part 3 Risk

~
The above example considered two overpriced stocks and a straight SML. Securities
Iying above the SML are underpriced. Their prices must rise until their expected
returns lie on the line. If the SML is itself curved, many stocks would be mispriced.
In equilibrium, all securities would be held only when prices changed so that the
SML became straight. In other words, linearity would be achieved.
4. The Capital-Asset-Pricing Model. You may remember from algebra
courses that a line can be described algebraically if one knows both its íntercept indii
and its slope. We can see from Figure 9.10 that the intercept of the SML is RF' for j
Because the expected return of any security with a beta of l is RM' the slope of the
line is RM - Rp This allows us to write the SML algebraically as secu
Capital-asset-pricing model:
R = RF + 13 (RM - RF) (9.18)
Difference between The
Expected Risk Beta
expected return we
return on free + of the x
on market and
a security rate security
risk-free rate

According to financial economists, the above algebraic formula describing the SML
is called the capital-asset-pricing model. The formula can be illustrated by assum-
ing a few special cases.

a. Assume that 13 = O. Here, R = RF' that is, the expected return on the
Be(
security is equal to the risk-free rate. We argued this in point (1) above.
tha
b. Assume 13 = 1. The equation reduces to R = RM' that is, the expected
return on the security is equal to the expected return on the market. the
We argued this in point (2) above. Thl
As with any line, the line represented by equation (9.18) has both a slope and ass
an intercept. RF' the risk-free rate, is the intercept. Because the beta of a security is Po
the horizontal axis, RM less RF is the slope. The line will be upward-slopíng as long lin
as the expected return on the market is greater than the risk-free rate. Because the Th
market portfolia is a risky asset, theory suggests that its expected return is above of
the risk-free rate. In addition, the empirical evidence of the previous "chapter
showed that the actual return on the market portfolia over the past 63 years was frc
well above the risk-free rate. Fi
Sť
• EXAMPLE p(
The stock of Aardvark Enterprises has a beta of 1.5 and that of Zebra Enterprises has Ol
a beta of 0.7. The risk-free rate is 7 percent and the difference between the
expected return on the market and the risk-free rate is 8.5 percent. l 7 The expected
returns on the two securities are
c
•
17 As reponed in Table 8.2, lbbotson and Sinquefield found that the expected return on common stocks was 12.1
•
percent over ) 926-1988. The average rísk-free rate over the same time interval was 3.6 percent. Thus, the average
difference between the two was 8.5 percent (12.)% - 3.6%). Financial economists use this as the best estímate of
•
the difference to occur in the future. We will use it frequemly in this text.
 Return and Risk Chapter 9 287

,It SML. Securities Expected return for Aardvark:

II their
expected 19.75% = 7% + 1.5 X 8.5% (9.18')
Id be mispriced.
I Expected return for Zebra: ~
nged so that the
p. 12.95% = 7% + 0.7 x 8.5%
from algebra 5. Portfolios as well as securities. Our discussion of the CAPM considered
th its intercept individual securities. Does the relationship in Figure 9.10 and equation (9.18) hold
f the SML is RF' for portfolios as well?
the slope of the Yes. To see this, consider a portfolio formed by investing equally in our two
securities, Aardvark and Zebra. The expected return on the portfolio is
Expected return on portfolio:
(9.18) 16.35% = 0.5 X 19.75% + 0.5 X 12.95% (9.19)
n
The beta of the portfolio is simply a weighted average of the two securities. Thus
we have
Beta of portfolio:
,cribing the SML 1.1 = 0.5 X 1.5 + 0.5 x 0.7
i
rated by assum- Under the CAPM, the expected return on the portfolio is
16.35% = 7% + 1.1 x 8.5% (9.20)
return on the
ooínt (1) above. Because the value in (9.19) is the same as the value in (9.20), the example shows
that the CAPM holds for portfolíos as well as for individual securities.
, the expected
the market.
6. Apotential confusion. Students often confuse the SML in Figure 9.10 with
the capital-market line (line II in Figure 9.8). Actually, the lines are quite different. ./
The capital-market line traces the efficient set of portfolios formed from both risky r,
DOtha slope and assets and the riskless asset. Each point on the line represents an entire portfolio.
a of a security is Point A is a portfolio composed entirely of risky assets. Every other point on the I
~sloPing as long line represents a portfolio of the securities in A combined with the riskless asset. I
are. Because the The axes on Figure 9.8 are expected return of aportfolio and the standard deviation
return is above of a portfolio. Individual securities do not lie along line II.
evious ~chaptec The SML in Figure 9.10 relates expected return to beta. Figure 9.10 differs
1St63 years was from Figure 9.8 in at least two ways. First, beta appears in the horizontal axis of ,
Figure 9.10 but standard deviation appears in the horizontal axis of Figure 9.8.
Second, the SML in Figure 9.10 !:.olds botlr for all individual securities and for all I
possible portfolíos, whereas line II (the capital-market line) in Figure 9.8 holds
Enterprises has
only for efficient portfolios.
:e between the
17 The expected

Concept Questions
• Why is the SML a straight line?
mon stocks was 12.1 • What is the capital-asset-pricing model?
t. Thus, the average
• What are the differences between the capital-market line and the security-
the best esli mate of
market line?
288 Part 3 Risk

In other
'.8 Summary and Condusions related t
This chapter sets forth the fundamentals of modern portfolio theory. Our basie
points are these:
1. The previous chapter showed us how to calculate the expected return and Key Te
variance for individual securities, and the covariance and correlation for Portfolio, 2
pairs of securities. Given these statistics, the expected return and variance for Opportuni
a portfolio of two securities A and B can be written as Efficient set
Expected return on portfolio = X)?A + X;RB Systematic
Var(portfolio) = X~<T~ + 2XAXB<TAB + X~<T~ Diversifiabl
(unsystem
2. In our notation, X stands for the proportion of a security in one's portfolia. Risk-averse,
By varyíng X, one can trace out the efficient set of portfolios. We graphed the Capital mar
efficient set for the two-asset case as a curve, pointing out that the degree of
curvature or be nd in the graph reflects the diversification effect: The lower
the correlation between the two securities, the greater the bend. Without
Suggest
proof, we stated that the general shape of the efficient set holds in a world
of manyassets. The capiu
3. Just as the formula for variance in the two-asset case is computed from a Sharpe,
of Risk."
2 x 2 matrix, the variance formula is computed from an NXN matrix in the
Lintner,J.
N-asset case. We show that, with a large number of assets, there are many
Finance
more covariance terms than variance terms in the matrix. In fact, the vari-
ance terms are effectively diversified away in a large portfolio but the
covariance terms are not. Thus, a diversified portfolio can only eliminate
some, but not all, of the risk of the individual securities.
4. The efficient set of risky assets we spoke of earlier can be combined with
riskless borrowing and lending. In this case, a rational investor will always
choose to hold the portfolio of risky securities represented by point A in Fig-
ure 9.8, then can either borrow or lend at the riskless rate to achieve any 9.2
desired point on the capital-market line.
5. If (1) all investors have homogeneous expectations and (2) all investors can
borrow and lend at the riskless rate, all investors will choose to hold the a.
portfolio of risky securities represented by point A. They will then either
borrow or lend at the riskless rate. In a world of homogeneous expectations,
point A represents the market portfolio. 9.3
6. The contribution of a security to the risk of a large portfolio is the sum of
the covariances of the security's return with the returns on the other se- a.
curities in the portfolio. The contribution of a security to the risk of the
market portfolio is the covariance of the security's return with the markeťs
return. This contribution, when standardízed, is calied the beta. The beta of a b.
security can also be interpreted as the tesponsiveness of a security's return
to that of the market.
7. The CAPMstates that
R = RF + I3(RM - RF)
 Return and Risk Chapter 9 289

In other words, the expected return on a security is positively (and linearly)

related to the security's beta.
falia theory. Our bas ic

xpected return and KeyTerms

id correlation for Portfolio, 256 Separation principie, 276
eturn and variance for Opportunity (feasible) set, 263 Homogeneous expectations, 277
Efficient set, 264 Market portfolio, 278
·XsRB Systematic (market) risk, 272 Characteristic line, 283
X~(J~
Diversifiable (unique) Security market line, 285
(unsystematic) risk, 272 Capital-asset-pricing model, 286
y in one's portfolio. Rísk-averse, 272
lios. We graph ed the Capital market line, 276 ,J
ut that the degree of
effect: The lower
e bend. Without
et holds in a world
SuggestedReadings
The capital-asset-pricing model was originally publisbed in these two classic articles.

mputed from a
XN matrix in the
l, there are many
: In fact, the vari-
folio but the
I only eliminate
Questions and Problems
: combined with 9.1
resror will always
by point A in Fig-
e to achieve any 9.2
D all investors can
ese to hold the
~i11then either
ieous expectations,
9.3
io is the sum of
'the other se-
e risk of the
(iththe markeťs
beta.The beta of a
securíty's return
Sharpe, W. F. "Capital Asset Prices: A Theory of Market Equilibrium under Conditions
of Risk." journal of Finance (September 1964).
Lintner, J. "Security Prices, Risk and Maximal Gains from Díversíflcatíon." journal of
Finance (December 1965).
A portfolío consists of 20 shares of Andrews stock which sells for $50 per
share and 30 shares of Dean stock which sells for $20 per share. What are
the weights of the two stocks in this portfolio?
Security F has an expected return of 10% and a standard deviation of 5%
per year. Security G has an expected return of 20% and a standard devia-
tion of 60% per year. The correlation between F and G is 0.5-
a. What is the expected return on a portfolio composed 40% of security F
and 60% of security G?
b. What is the standard deviation of this portfolio?
Suppose the expected returns and. variances of stocks A and B are RA
0.2, RB = 0.3, <TÄ = 0.1, and <T~ = 0.2, respectively.
a. Calculate the expected return and variance of a portfolio that is corn-
posed of 60% A and 40% B when the correlation coefficient between
the stocks is - 0.5-
b. Calculate the expected return and variance of a portfolio that is com-
posed of 60% A and 40% B when the correlation coefficient between
the stocks is - 0.6.
C. How does the correlation coefficient affect the variance of the
portfolio?